The Harding Mine, Taos County, New Mexico

New Mexico Geol. Soc. Guidebook, 27th Field Conf., Vermejo Park, 1976 263 THE HARDING MINE TAOS COUNTY NEW MEXICO

RICHARD H. JAHNS Department of Geology Stanford University Stanford, California RODNEY C. EWING Department of Geology The University of New Mexico Albuquerque, New Mexico

INTRODUCTION collecting on a modest scale. Anyone with an interest in visit- The Harding mine, in the western part of the Picuris Range ing the property should contact the Chairman, Department of about 20 miles southwest of Taos (Fig. 1), has yielded substan- Geology, University of New Mexico, Albuquerque, New tial amounts of commercial beryl, lepidolite, spodumene, and Mexico 87131. A splendid collection of representative Harding tantalum-niobium minerals over a period of half a century. It minerals, donated by Dr. Montgomery, can be viewed in the also has become widely known as a source of handsome University's Geology Museum. mineral specimens, as a provocative locality for scientific The published record of the Harding mine comprises early studies, and as an attraction for those who appreciate spec- general descriptions (Roos, 1926; Just, 1937) and mineralogic tacular exposures of pegmatite. It lies four miles southeast of notes (e.g., Schaller and Henderson, 1926; Hirschi, 1928, the Rio Grande Canyon at an altitude of 7400 ft (sec. 29, T. 1931; Hess, 1933), summaries of extensive exploratory work 23 N., R. 11 E.), and it is readily accessible from a nearby by the U.S. Bureau of Mines in 1943 and 1948 (Soule, 1946; point on State Highway 75 about 6 miles east of Dixon. Berliner, 1949), descriptions of milling operations for tan- The mine property has been generously leased by the talum minerals (Wood, 1946) and mining operations for beryl owner, Dr. Arthur Montgomery, to the University of New (Montgomery, 1951), and various discussions of mineralogic Mexico for preservation as one of the State's unusual natural and structural relationships in the pegmatite bodies (e.g., assets. The University plans also to make this classic locality continually available for public inspection, study, and mineral

264 JAHNS and EWING Cameron and others, 1949; Montgomery, 1950; Page, 1950; of J. L. Danziger of Los Angeles, began construction of a Jahns, 1951; 1953a,b; Mrose, 1952; Rimal, 1962). The pur- crushing and grinding plant near the railroad depot at pose of the present paper is briefly to describe the mining Embudo, and in 1927, after the mill had been completed, this history and geology of the Harding locality, in part as an aid to company also assumed control of new work at the mine. The those who visit the property or who view some of its charac- main pit was much enlarged, and as it merged with the east cut teristic minerals in Albuquerque at the Geology Museum. The and grew toward its present elongate form (Fig. 2), a second treatment is based on detailed field studies that began in 1942 adit was driven to it from the west. Both this adit and the under the aegis of the U.S. Geological Survey and continued main adit ultimately were daylighted, and they now appear as intermittently for more than two decades, and on mineralog- main and west entries to the quarry. The hand-sorted ore, ical investigations in several laboratories. broken at the mine to a maximum dimension of 9 inches, was We are indebted to John W. Adams and William P. Irwin of trucked to the Embudo mill where it was passed through a jaw the U.S. Geological Survey and to Lauren A. Wright of The Pennsylvania State University for important contributions during the early periods of field work, and to Arthur Mont- gomery of Lafayette College and Lincoln R. Page and the late Waldemar T. Schaller of the U.S. Geological Survey for numerous penetrating discussions of the Harding mineral associations. We wish also to acknowledge the valued aid of Stuart A. Northrop of the University of New Mexico in com- piling an authoritative list of minerals found in the Harding pegmatites. It is a special pleasure to recall, in connection with the field investigations, the hospitality and many courtesies received from Arthur Montgomery, John H. Soule of the U.S. Bureau of Mines, and Eliseo Griego, the late Flaudio Griego, and Lydia and "Doc" Zellers of Dixon.

HISTORICAL SKETCH The Lepidolite Chapter, 1919-1930 Scattered large blocks and low but prominent outcrops of pegmatite quartz attracted prospectors to the Harding locality before 1900 but it was not until 1918 that Joseph J. Peyer of Taos recognized accompanying large, compact masses of lilac- colored lepidolite as indications of significant lithium mineralization. Beginning in 1919 Peyer and two partners, Frank Gallup of Taos and Arthur H. Gossett of Embudo, broke up the tough, bouldery masses by means of TNT-rich poultices, and obtained additional lepidolite from the main pegmatite ledge by open-cut work on the north-facing hillside. They hand-sorted the broken ore, hauled it by wagon to Embudo via a narrow access road they blasted across the high ground east of Dixon, and thence shipped it by rail to Wheeling, West Virginia, for grinding and sale to the ceramic industry. At that time the material was used mainly in the manufacture of opaque white glass for jar tops and indirect lighting fixtures, as it was an effective flux, opacifier, and toughener. Late in 1920, after parts of the deposit had been explored by shot-core drilling, operations were taken over by the Mineral Mining and Milling Corporation of New York under lease and assignment agreements. An adit was driven southward into the hillside to permit expansion of the open cut by glory-hole methods, and a second cut, adjacent on the east, later was opened from a short entry (northeast entry, Fig. 2). By 1923 production of rich shipping ore had reached a level of nearly 800 tons per year. The economics of the enterprise proved to be disappointing, however, and mining was sus- pended in 1924. During that year two shipments of apparently excellent lepidolite caused grave difficulties when the material was used for making glass. Sketchy reports indicate that this material may have contained substantial quantities of the tantalum mineral microlite; indeed, some of it may have been high-grade Ta-Nb ore. In 1924 the Embudo Milling Company, under the direction

crusher and then continuously ground in a quartzite-lined mill. accessible, was developed opposite the main quarry entry (Fig. Six-inch lumps of ore served as the principal grinding agent. 4). The mine was closed in 1930, and the property transferred After screening and passage through a magnetic separator, the back to the original owners in 1931. product was sacked for direct shipment to glass manufacturers. The score sheet for the opening chapter of lepidolite opera- There was considerable demand for the ground lepidolite, tions, 1919-1930, shows mixed results, among them (1) ship- especially in the production of glasses resistant to thermal ments of broken ore and ground products amounting to shock, and the output soon reached levels far beyond those of approximately 13,500 tons, (2) a gross output value of about the earlier years. But expenses also were great, and in 1928, $140,000, f.o.b. Embudo, (3) small net profits for three years for example, a net income of only $619.42 was generated but a net loss for the entire period of mining, and (4) develop- from shipments of 3612 tons valued at $58,858.00. Quality of ment of a large quarry, appended openings, and many the product was maintained at high levels, and it was not scattered small workings that together provide some of the difficult to meet the specified tolerance of 0.1 percent for iron world's finest exposures of mineralogically complex pegmatite. content. However, much of the output contained both lepidolite and lithian muscovite, so that the usual tenor requirement The Microlite Chapter, 1942-1947 of 3.0 percent of Li20 was not so easily satisfied unless some The Harding mine was reborn in 1942, when the United spodumene was included with the lepidolite. Undoubtedly States was in such critical need of tantalum metal that an most shipments contained a fraction of spodumene, which was almost frantic search was under way for domestic sources of closely associated with the micas in the mined ore. ore. Microlite, a relatively rare calcium tantalate, had been By 1929 the large, pod-like body of rich lepidolite ore had reported from the main quarry (e.g., Hirschi, 1931) and was been essentially mined out, and efforts were turned increas- known in a few rock specimens collected there during earlier ingly toward the more expensive extraction of irregular projec- mining. Small amounts of tantalite-columbite also had been tions and shoots. Several undercuts were made from the south recovered from localized placer deposits north and west of the face of the quarry (Fig. 6), and finally a large opening, still mine, but the property hardly was regarded as a potentially

267

detail. This work confirmed Montgomery's favorable appraisal, and late in 1942 he entered a lease-purchase agreement for the property. With a crew of local miners, he sorted and sacked five tons of dump material containing more than 10 percent of microlite. He soon turned his attention to the part of the mine whence this material had come, ground that presumably also was the source of the unsatisfactory lepidolite shipped for

glass making in 1924. In Montgomery's own words, "The mining that ensued along the low east wall of the quarry surely

was one of the strangest ever seen. Six men attacked that rock wall with chisels. The ore was far too valuable to be blasted. And it proved possible to chisel, pry, and hammer it out of the surrounding spodumene-rich rock by hand. Primitive mining, yes, but astonishing results were forthcoming. The small lepidolite ore masses showing in the wall grew bigger and bigger as they became fully exposed.... It was incredibly rich ore,

quite as valuable as high-grade ore in economic importance, and it came out of the quarry wall in amazing quantity."

Sorting of the ore was a straightforward but difficult proce- dure, as much of the microlite occurred in the form of tiny, pale yellow to almost colorless grains within irregular aggregates of micas (Fig. 3) and was easily overlooked. Montgomery soon developed considerable expertise in the rapid identification of high-grade material by means of two simple tests based on physical properties of the microlite. Its high density gave an unusual heft to rich specimens, and its very high index of refraction clearly revealed the outlines of individual crystals on

saliva-moistened rock surfaces. The sorted ore, though rich, required mechanical concentration that no commercial firm could provide for small lots of such unusual and valuable material. Fortunately a special arrangement was made for use of the U.S. Bureau of Mines Testing Laboratory in Rolla, Missouri for the experimental processing of 33.5 tons of ore by gravity methods. The result-

ing yield, obtained by mid-1943, was by far the largest single unit of domestic tantalum production ever recorded-6137 pounds of concentrate containing 71.14 percent of Ta2 05 and 6.41 percent of Nb205. During the following two years several smaller lots of ore were batch-milled, first in a facility at the California Institute of Technology and later at commercial ore- testing stations. In 1945 a small mill was built near the Rio Grande at Rinconada, in order directly to serve the mine through recovery of microlite and tantalite-columbite from a

mixed feed of high- and low-grade materials. This facility, which embodied hammermill grinding followed by gravity

separation in Humphreys spiral launders (Wood, 1946), re- portedly yielded fair to good recoveries but fell short of its rated capacity of 12 tons per 8-hour day. In response to a special request by the War Production Board, a carload of high-grade spodumene was produced in 1943, mainly from the eastern part of the quarry. This material was hand-picked from pegmatite containing relatively important source of tantalum when Arthur Montgomery large, lath-like crystals (Fig. 6), a very tedious process because visited it to check for occurrences of microlite. He soon found these crystals were so thin and fragile that they were separated rich material in one of the old dumps (Fig. 4), as well as small into small fragments by blasting. In 1946, some lepidolite ore concentrations of similar material in place near the east end of of good quality also was recovered during the mining opera- the quarry (Fig. 2). tions for microlite. By the time production of tantalum mine- The U.S. Geological Survey then was called upon to sample rals ceased in 1947, the eastern parts of the main quarry face the microlite-bearing dump and to map the mine workings in had been considerably reshaped, and a labyrinth of tunnels

and small, finger-like rooms had been developed south of this face (Fig. 4). Most of the underground workings have since become inaccessible or unsafe for entry. JAHNS and EWING

nearly 100 pounds, and he soon recovered about 4 tons of smaller chunks from nearby parts of this dump. Then turning his attention to a small knob exposed high on the quarry face

near the west entry, he removed the overlying country rock to expose a thick, elongate lens of nearly pure beryl lying at the upper surface of the pegmatite "much like a thick dab of frosting on the top of a cake." The 23-ton shipment from this occurrence contained nearly 11 percent of Be0.

The 1943 results of U.S. Bureau of Mines exploration also provided encouragement. Several of the drill holes had penetrated beryl in various parts of the main pegmatite body, and in one of them (No. 22, Fig. 4), about five feet of nearly pure beryl in many parts of this hanging-wall zone was demonstrated by detailed surface mapping and, on the hill- slope immediately west of the quarry, by 26 sample trenches with an aggregate length of 788 ft (Soule, 1946). The Harding was born as a full-scale beryl mine in 1949, when Montgomery and Flaudio Griego of Dixon began to drive a tunnel southwestward toward USBM Drill Hole 22 from a point in the quarry near the previously removed lens of

beryl. After initially disappointing results, the new activities began to yield remarkable returns as increasingly larger and more continuous masses of ore were encountered. By early 1950 the working faces of the tunnel and some of its branches were in solid beryl for days at a time, and production of

hand-sorted material rose to levels of more than a ton per day. The sorting, no easy task whether done in daylight or under- During 1943 the U.S. Bureau of Mines explored south of ground by the light of a Coleman lantern, was based in part on the quarry by means of 39 diamond-drill holes with an aggre- two distinctive features of the beryl, a faintly greasy to resin- gate length of 4371 ft, and in 1948 seven additional explor- ous luster and a rude basal cleavage along which the coarsest atory holes totaling 1 1399 ft were drilled. The recovered cores material typically broke into thick plates and slabs. were logged by the U.S. Geological Survey, and assayed by the Shipments reached the 150-ton level in both 1950 and Bureau of Mines. These investigations led to definition of a 1951, when they placed New Mexico in the forefront of all considerable tonnage of pegmatite with sufficient content of beryl-producing states, and they averaged nearly 100 tons per spodumene and tantalum-niobium minerals to be classified as year from 1950 until operations ceased with Griego's death in potential milling-grade ore. Thus the Harding became the first 1958. Yet these activities were extremely modest by normal domestic pegmatite mine with blocked-out reserves, that mining standards. At no time was a crew of more than four remain in place at the present time. men used, and no mechanical concentration ever was involved. The unusual microlite operations and allied activities of the Nearly all mine haulage was done by a sturdy $12.00 he-mule five-year period 1942-1947 were relatively small in scale, who answered to the particularly appropriate name Beryl. simple in style, and prevailingly successful. Among the results The impressive score for the 1950-1958 beryl chapter of were (1) a prompt and effective response to urgent wartime Harding mine history includes (1) production of 690 tons of needs for tantalum and lithium, (2) a yield of 41 tons of concentrates with an average Be0 content of 11.2 percent, as high-grade spodumene, 558 tons of lepidolite ore, nearly 500 well as 184 tons of lower-grade material with an estimated pounds of placer tantalite-columbite containing an average of average Be0 content of 5.5 percent, (2) incidental small pro-

43 percent Ta205, and 22,116 pounds of microlite concen- duction of lepidolite ore, mainly in 1950 and 1951, (3) a trates containing an average of 68 percent Ta2O5 and 7 highly profitable operation from which a large fraction of the percent Nb205, and (3) sufficient net income from produc- returns was assigned to the needs and interests of residents in tion to permit purchase of the property. the Dixon area, and (4) further proof that the spatial distribution of many economically desirable pegmatite minerals, The Beryl Chapter, 1950-1958 however irregular, is nonetheless systematic (e.g., Cameron and Beryl was identified at the Harding locality during the early others, 1949). years of mining (e.g., Just, 1937, p. 26), but it was regarded as no more than a rare accessory mineral in the pegmatite. That it GEOLOGIC SETTING actually is a widespread and locally abundant constituent, easily mistaken for quartz or coarse feldspar, was initially The Harding pegmatites lie within a complex terrane of Pre- cambrian rocks that have been mapped, described, and recognized by the U.S. Geological Survey late in 1942. Its variously interpreted by Just (1937), Montgomery (1953), and economic potential at first was viewed in terms of recovery by Long (1974). This terrane is characterized by a thick section milling, but Arthur Montgomery subsequently discovered very of arenaceous and pelitic metasedimentary rocks with inter- coarse material readily amenable to hand-cobbing. In 1944 he layered metavolcanic rocks of basaltic to rhyolitic composi- found, on one of the old dumps, a block of beryl weighing tion, and by large intrusive bodies of younger igneous rocks that have been collectively referred to as the Dixon Granite or

HARDING MINE 269

Embudo Granite. Both Just and Montgomery recognized and dip southward at low to moderate angles. Several thinner, important textural and compositional differences among the steeply dipping branches and connections trend north to exposed granitic rocks, and more recently Long (1974) has northwest. The main body, at the western end of the belt, is shown that they probably represent at least four distinct well exposed in outcrop and mine workings along the south- episodes of magmatism occurring at progressively increasing western side of a shallow canyon (Fig. 2), and an extensive depth over a considerable span of Precambrian time. up-dip erosional remnant lies opposite on the flank of North The east-west outcrop belt of pegmatite bodies at the Hard- Knob. The ridge extending east-southeastward from this knob ing locality corresponds in general position with a complicated is marked by the outcrops of several thinner dikes, most of boundary between quartz-muscovite schists on the north and which have been opened by means of pits, cuts, and short amphibolitic rocks on the south (Fig. 1). The well-defined tunnels. The easternmost of these bodies can be traced across a planar structure in these rocks trends northeast to east-north- steep-walled canyon that drains southward into the Rio de east and dips steeply southeastward (Figs. 1, 2), and the Penasco. boundary between them has a similar general attitude in both All the dikes are granitic in bulk composition and consist directions beyond the pegmatite area. No granitic rocks are chiefly of quartz, potash feldspar, and albite. Muscovite is exposed in the immediate vicinity of the mine, but less than a widespread, along with lesser amounts of accessory apatite, mile to the east the prominent Cerro Alto consists of intrusive beryl, garnet, magnetite, microlite, and tantalite-columbite. metadacite, and in the area west of the mine porphyritic Some of the bodies are grossly homogeneous or contain no quartz monzonite underlies Cerro de los Arboles and Cerro more than pods of quartz and other local expressions of Puntiagudo. To the south, in the canyon area traversed by the internal zoning, but those in the central and western parts of Rio de Penasco (Rio Pueblo), are many exposures of biotite the belt comprise extensive and well-defined internal units of quartz monzonite and granodiorite that represent a large markedly contrasting lithology. It is in these distinctly zoned pluton of considerable complexity (Long, 1974). bodies that nearly all major concentrations of beryl, lepidolite, Bodies of pegmatite, youngest among the Precambrian rock spodumene, microlite, and tantalite-columbite are found. units, are widespread in the region and are especially abundant Some of the Harding dikes are enclosed by amphibolitic in the belt of metamorphic rocks traversed by State Highway rocks, chiefly hornblende-plagioclase schist with layers and 75. Many are long, thin dikes, with northeasterly trend and pods of epidosite and dark, impure quartzite; other dikes steep dip, that consist of quartz, alkali feldspars, muscovite, extend northward into the terrane of quartz-muscovite schists garnet, and sparse yellowish green to green beryl in well- (Fig. 1). The main dike is overlain at the present outcrop level formed prisms. Most of these dikes are internally zoned, with by the amphibolitic sequence (Fig. 2) and underlain by schists prominent core segments of massive quartz. Other zoned with numerous thick layers of amphibolite. Alkali meta- bodies, in general smaller and more pod-like, consist almost somatism of the wall rocks adjacent to pegmatite is indicated wholly of quartz, albite, and muscovite. Bodies of a third kind, by extensive conversion of hornblende to biotite, and by characterized by greater thickness, prevailingly gentle dip, and abundant metacrysts of potash feldspar, muscovite, and albite. unusual abundances of Be, Li, and Ta-Nb minerals, are known Some quartzite and quartz-rich schist on North Knob is only in the mine area and at a locality about half a mile to the heavily impregnated with lithium-bearing micas. northwest. Marked differences in composition, structure, and host-rock relationships suggest that the several kinds of pegmatite in the The Main Dike region may be of different ages, but a sequence cannot be General Relationships firmly established and dated from radiometric information The main Harding dike is highly irregular in surface plan, now available. Aldrich and others (1958) reported K-Ar and with an exposed length of about 1100 ft and breadth ranging Rb-Sr ages of 1260 m.y. B.P. for four Harding muscovites and from a few feet to as much as 250 ft in the quarry area (Fig. 1350 m.y. B.P. for a Harding lepidolite, and Gresens (1975) 1); maximum breadth is considerably greater if the pegmatite reported a K-Ar age of 1335 m.y. for muscovite from an on North Knob is included. The principal quarry face (Fig. 2), unnamed pegmatite in the Picuris Range. Neither these ages trending west-northwest approximately parallel to the original nor that suggested by Fullagar and Shiver (1973) for granitic outcrop, provides more of a cross section than a longitudinal rocks in the region can be unequivocally translated into a section, because in three dimensions the body has the general dated episode or sequence of plutonism, but Long (1974) has shape of a wide and nearly flat, southwestward-extending pointed out that the Harding pegmatites may be as much as tongue with tapering, sharply downturned lateral margins and 100 million years younger than the youngest of the granitic a central region SO to 80 ft thick. The gently undulatory major plutonites. An age difference of such magnitude is consonant axis of the tongue has an average plunge of about 10 degrees with field and photographic evidence suggesting that these and an explored length of nearly 1600 ft. The original length is pegmatites may not be genetically related to any of the not known, as up-plunge parts of the body have been removed granitic intrusives now exposed in the Picuris Range. by erosion beyond North Knob. Upper and lower surfaces of the tongue are broadly wavy THE PEGMATITE BODIES (Fig. 2), and in places they also are marked by sharp corruga- General Features tions with amplitudes of a few inches to at least 15 ft. Nearly The Harding pegmatites crop out in a well-defined belt all these rolls plunge at low angles, and in general their trend is about 2500 ft long and 150 to 500 ft wide (Fig. 1). Most of parallel to the strike of foliation in the adjacent country rocks. the bodies are pinching-and-swelling dikes that in general are Parallel to this foliation in both strike and dip is a well-defined 5 ft to 20 ft or more in thickness, trend west to west- set of numerous fractures that transect the pegmatite. A fault northwest, with like attitude and vertical separation of 5 to 8 ft also cuts

JAHNS and EWING

of 20 to 35 ft. Muscovite, microcline, and albite of the cleavelandite variety are present locally, and both cleavelandite and muscovite are widespread and abundant constituents of an

analogous massive-quartz unit in the footwall half of the dike (Fig. 5). The upper massive-quartz zone is underlain by a spectacular quartz, lath-spodumene zone. This widespread unit is 6 to 20 ft thick in surface exposures, and reaches subsurface thicknesses

of 20 to 40 ft at places where it fills the entire interior of the dike. The spodumene characteristically occurs as long, thin, blade-like crystals of moderate to great size, usually arranged in jackstraw fashion. In the eastern half of the main quarry, the crystals are oriented somewhat more regularly to form a

crudely comb-like pattern (Figs. 2, 6). Beryl, apatite, white to green microcline, and Ta-Nb minerals are present locally in the lower part of this zone. Some portions of the quartz, lath- spodumene rock have been pseudomorphously replaced, the spodumene selectively by rose muscovite and the quartz by

cleavelandite. Similar replacement rock is present in footwall parts of the dike (Fig. 5). The core of the dike is distinguished by varieties of lithium- and tantalum-bearing pegmatite that in general are less coarse grained than rocks of the overlying zones. The dominant

variety, known as "spotted rock," is a relatively even-grained aggregate of spodumene, microcline, and quartz, with various combinations of accompanying finer-grained albite, lithium- bearing muscovite, lepidolite, microlite, and tantalite- columbite. Much of the spodumene and potash feldspar occurs the pegmatite exposed near the inner end of the west entry as rounded masses about an inch in average diameter, with and the outer end of the main entry (Fig. 4). surfaces that are ragged in detail (Fig. 7). Both minerals have Internal Zoning been extensively corroded and replaced by micas in aggregates The main dike displays a remarkable segregation of constitu- of tiny flakes. ent minerals into contrasting zones whose general shape and The "spotted rock" contains an average of 12 to 15 percent distribution reflect the upper and lower surfaces of the body of spodumene and about 0.15 percent of Ta-Nb minerals, and (Fig. 5). Most of the zones thus appear as wide subhorizontal hence might be viewed as low-grade ore. The principal body of stripes on the main quarry face and in areas of extensive out- this rock has the general form of a loaf of French bread, with crop, although their surface distribution on North Knob is one prominent constriction and an equally prominent vertical complicated by the geometry of dip-slope exposure. For bulge (Figs. 4, 5). It is 20 to 55 ft in maximum thickness and present purposes, a brief description of zones in the thickest 50 to 175 ft wide (Fig. 4), with a long, or down-plunge extent portion of the main dike, well known in three dimensions of nearly 700 ft southwestward beyond the main quarry face. from surface exposures and exploratory drilling, will suffice to Its position is reflected in some places by a broad bulge in the illustrate the major aspects of segregation. hanging wall of the pegmatite body, and in others by a sag in The upper contact with country rocks is marked by a white the footwall (Fig. 5). border rind, 1/2 to 1 in. thick, of fine-grained quartz-albite- Typical "spotted rock" grades into lepidolite ore that con- muscovite pegmatite. Beneath this rind is a continuous layer, tains rounded masses of spodumene but relatively little potash typically 1 to 5 ft and locally as much as 10 to 14 ft thick, feldspar, and with decreasing spodumene content into bodies that also is rich in quartz, albite, and muscovite but is in large of nearly pure lepidolite. Several of these bodies were en- part very coarse grained. Potash feldspar is locally abundant, countered long ago in the main quarry, but little of the and apatite, beryl, and tantalite-columbite are widespread material remains in place today. Another core variant, accessory constituents. Spectacular concentrations of coarse to common in the eastern part of the quarry and also found very coarse beryl also have been found, especially beneath the locally in the western part, is distinguished by platy or lath- crests of major rolls that are most abundant in lateral parts of like individuals of spodumene 2 to 30 in. long. These occur the tongue. An analogous layer, characterized by the same either as isolated units (Fig. 7) or as jackstraw aggregates (Fig. accessory minerals but in general with less beryl, is present 8) with finer-grained matrices of spodumene, lepidolite, along the lower surface of the tongue. It is rich in coarse Li-muscovite, albite, quartz, and Ta-Nb minerals. Veinlets and quartz, but in many places its prevailing texture is more aplitic small, irregular masses of dark smoky quartz are present than pegmatitic and is distinctively sugary in appearance locally. This rock has been the host for most of the high-grade because of abundant fine-grained albite. tantalum ore thus far encountered, and it has been referred to Beneath the upper beryl zone is a continuous layer of as the largest known body of microlite in the world (Baker, massive quartz, 2 to 8 ft thick, that is well exposed in the main 1945). quarry and on the low knob immediately to the east. In some Bulk Composition down-plunge parts of the dike it reaches subsurface thicknesses The main pegmatite body, as exposed in the quarry and

HARDING MINE 271 penetrated by drill holes, has been extensively sampled and matites and materials collected from them. Its purpose is to analyzed (Table 1). The pegmatite is typically granitic in terms provide notes on the most abundant or important species at of major constituents, with a high Na2 0/K2 0 ratio that in this locality, and to list the remainder as additional search part is reflected by an abundance of albite in footwall parts of targets for the diligent collector (Table 5). Reports by other the body that are not well exposed. It is low in iron and investigators are referenced for special features or for species strikingly deficient in calcium and magnesium. Remarkable for we have not personally seen and confirmed. For many useful any large mass of granitic rock are its high contents of lithium details not included here, reference should be made to (mainly in spodumene and micas), cesium and rubidium Minerals of New Mexico (Northrop, 1959). (mainly in micas, potash feldspar, and beryl), and fluorine (mainly in micas, apatite, and Ta-Nb minerals). It also contains Major Minerals about 0.03 percent of Be and 0.08 percent of Ta + Nb. Quartz [Si02 By far the most widespread pegmatite The composition of the border rind along the hanging wall mineral. Occurs prominently as thick layers and pod-like (column 2, Table 1) is distinctly different from that of the masses of milky white to smoky gray anhedral crystals 0.5 in. pegmatite body as a whole, hence the rind cannot be readily to 3 ft in diameter, some with rude rhombohedral cleavage, viewed as a conventional chilled margin. It is relatively rich in and as very coarse aggregates interstitial to large crystals of Na20 and poor in K20 and Si02, and it may well be composi- spodumene (Fig. 6). Also forms extensive fine-grained, sugary- tionally complementary to added constituents in the zone of appearing aggregates with albite and microcline. Common else- potash and silica metasomatism in adjoining country rocks. where as a fine- to medium-grained interstitial or groundmass The Si0 content of "spotted rock" (third column of Table 2 mineral. Light to very dark smoky quartz forms scattered 1) corresponds to a relatively low percentage of quartz, the irregular veinlets and small pods, in places containing well- high K 0/Na 0 ratio to an abundance of potash feldspar and 2 2 faced crystals of beryl, spodumene, uranian microlite- micas, and the high Li 0 content to an abundance of spo- 2 pyrochlore, and tantalite-columbite, also rare allanite, thorite, dumene and lithium-bearing micas. Listed for comparison is an and zircon. analysis for a section through the pegmatite as provided by the Albite [NaAlSi308]. Occurs as fine-grained aggregates, south face of the quarry. "Spotted rock" is well represented in especially with quartz and muscovite in outer parts of dikes, this section, hence it is not surprising that the analytical results and also as the variety cleavelandite in lustrous curving plates are intermediate between those listed in the first and third and blades 0.3 to 2.5 in. long. Common as coarse fracture- columns of the table. filling and replacement aggregates in quartz. Generally white, MINERALOGY also light gray and very pale green. Compositional range Ab91-Ab99, with the remainder mainly or entirely Or; most The following brief summary of Harding minerals and their coarse material is nearly pure Ab. occurrence is based chiefly on our own studies of the peg- Microcline [KAISi308] . Widespread as extensive aggregates of coarsely perthitic subhedral crystals 1 in. to 4 ft in diameter, and as large, roughly euhedral individuals in massive quartz. Typically cream-colored, gray, and pale pinkish tan. Also abundant with quartz and muscovite in fine-grained aggregates. Locally interstitial to large laths of spodumene as white to light green (amazonstone) crystals 1 to 3 in. in diameter. A major constituent of "spotted rock" as 1/2- to 4-in. individuals with rounded and corroded margins (Fig. 7); non-perthitic and in general almost soda-free, ranging from white to lilac according to degree of impregnation by extremely fine-grained lepidolite and Li-muscovite. Some pale pink masses may have been mistaken for amblygonite by early observers (e.g., Hirschi, 1931).

Muscovite [KAl2(AlSi3)010 (OH)2] I. Abundant in smaller dikes and in outer parts of main dike as pale yellow-green to green flakes and thick plates 0.1 to 1.5 in. in diameter,

typically containing 0.3 to 0.8 percent of Li20. In inner parts of several dikes as aggregates of tiny lilac, lavender, violet, pink, or gray flakes and plates forming large irregular pods or smaller masses interstitial to crystals of spodumene, cleavelandite, and other minerals. Such aggregates, somewhat translucent and waxy in appearance, generally contain 1.0 to 3.4

percent of Li2 0. Also occurs as columnar crystals 0.2 to 2.5 in. long (parallel to c-axis), typically in radiating aggregates with cleavelandite and quartz. A widespread rose variety, in bright, pink flakes 0.1 to 0.8 in. in diameter and generally with

less than 0.3 percent of Li20, occurs variously as fracture fillings and replacement masses in "spotted rock" and adjacent pegmatite, as overgrowths on crystals and crystalline aggregates of lepidolite, as pseudomorphs after spodumene, micro- 272 JAHNS and EWING cline, and topaz, and as complex intergrowths with lepidolites are slightly more bluish and are readily fusible on cleave-I andite. Gray to faintly lavender-gray Li-muscovite thin edges exposed to a hot flame. Firm identification also impregnates quartzose wall rocks in several places along ordinarily requires chemical analysis, optical studies under the main dike. microscope, or X-ray determination of crystal structure. See Characteristic variations in appearance and alkali composi- also Levinson (1953) and Rimal (1-962). tion among Harding muscovites are indicated in Table 2. See Spodumene [Li AlSi206 I. Typically white to very light also Roos (1926), Schaller and Henderson (1926), Heinrich gray, but faintly greenish or yellowish on many freshly broken and Levinson (1953), and Rimal (1962). surfaces. Abundant in massive quartz as thin, slightly tapering, Lepidolite [K(Li,Al)3 (Si,Al)401 0(F ,OH)2]. Most abun-dant lath-shaped crystals and crystal fragments a few inches to in "spotted rock" and other interior zones of main dike as several feet long, arranged in various parallel, radiating, and finely crystalline to extremely fine-grained translucent, crisscrossing patterns (Figs. 2, 6). Largest observed individual waxy-appearing masses 0.2 in. to tens of ft in maximum 14 ft long, 4 to 11 in. wide, and 0.5 to 0.9 in. thick. Most dimension. Color bluish gray, wine red, lilac, lavender, and crystals broken along cross-fractures, with fissure fillings of pale to deep purple, with general range of 3.5 to 4.6 percent quartz (Fig. 6) and rarely microcline. Also abundant in contained Li20. Also forms fracture fillings, embayments, and "spotted rock" as discoidal crystal fragments, 0.5 to 3 in. in replacement masses in microcline, less commonly in spodu- diameter, with irregular but grossly rounded margins, and in mene. Occurs with quartz and albite in several dikes and outer adjacent lithium-rich parts of main dike as jackstraw aggregates parts of main dike as individual thick crystals and as aggregates of laths 2 to 30 in. long (Fig. 8). In a few pods and veins of of lilac to pale purple flakes and plates 0.1 to 0.3 in. in dark smoky quartz as small, pale gray-green euhedral crystals. diameter. Some larger plates, 0.2 to 0.7 in. across, comprise Most crystals in interior parts of dikes are corroded, veined, several smaller crystals, each elongated parallel to its c-axis and and partly replaced by various aggregates of lepidolite, rotated about this axis relative to its neighbor, giving the three- Li-muscovite, rose muscovite, albite, and quartz. Some also are dimensional appearance of single tapering crystals or narrow partly altered to very fine-grained eucryptite, micas, and sprays of slender prisms. albite. Li20 content of relatively fresh crystals ranges from 6.0 Characteristic variations in appearance and alkali composi- to nearly 7.2 percent, with average of about 6.8 percent. tions among Harding lepidolites are indicated in Table 2. Distinction from the muscovites often is difficult, as many of Principal Accessory Minerals the properties and associations are similar, but in general the Apatite [Ca5 (PO4)3(F,OH,Cl)]. Irregular anhedral masses

HARDING MINE 273 and stubby, rough-faced crystals 0.2 to 5 in. in maximum prisms and anhedral crystals within small cleft-like cavities, dimension. Dark greenish blue to very deep blue, and strongly where associated with crystals of quartz, schorlite, epidote, fluorescent yellow-gray under ultraviolet light. Some crystals garnet, and magnetite. markedly color zoned, with blue interiors and gray, non-fluo- Variations in form, color, index of refraction, and some rescent rinds. Occurs mainly near walls of dikes, and locally in aspects of composition among Harding beryls are indicated in quartz-lath spodumene pegmatite. Table 3. Corresponding to occurrence from the walls inward to Beryl [Be3Al2Si6018]. Widespread in nearly all dikes, but by the core of the main dike are shifts in crystal habit from far most abundant along and near walls of main dike as prismatic to equant to tabular, a general increase in co index of equant anhedral to subhedral crystals 0.5 in. to 5 ft across, in refraction, a general decrease in Be0 content, and general places forming coarse lens- and layer-like aggregates of consid- increases in contents of Li20, Na2O, and Cs20. Common erable tonnage. Rude basal cleavage locally prominent. Some low-alkali beryl apparently occurs only in thin dikes with no large crystals with rough basal, prism, and pyramidal faces, and conspicuous lithium mineralization. The blue beryl from small a few noted by Arthur Montgomery as enclosing prismatic cavities in the country rocks is unusual in its high content of phantoms 3 to 4 in. long and 0.2 to 0.5 in. in diameter. Milky both Be0 and alkalies, and it may be Li-poor and Fe-rich like white, gray, and faintly pinkish, greenish, or yellow-greenish, the variety reported from Arizona by Schaller and others typically with resinous or greasy luster. Interior parts of some (1962).

crystals pale rose and transparent, with vitreous luster and Garnets [(Mn,Fe)3 Al2 (SiO4)3 ]. Wine red almandine as conchoidal fracture. Also occurs in interior parts of quartz, lath- small, widely scattered crystals. Locally abundant in fine- spodumene zone as white to pink equant crystals 0.3 to 4 in. grained footwall rocks of main dike. Pale salmon pink to in diameter; anhedral to subhedral and in part transparent. orange-red spessartine sparse to abundant in interior parts of Locally abundant as small, white to pale pink and yellowish lithium-bearing dikes, and common in albite-rich pegmatite of gray tablets with hexagonal outline in light to dark smoky the main dike as rough dodecahedral crystals 0.05 to 1.2 in. in quartz of core pegmatite, especially in eastern parts of main diameter. Larger crystals much fractured, and some heavily dike. Present sparingly in at least three other dikes as pale stained by manganese oxides. yellowish green stubby prismatic crystals 0.5 to 2 in. long. Microlite-pyrochlore [(Na,Ca,U)2 (Ta,Nb)2 O6 (O,OH,F)] . Locally in adjacent country rocks as blue to greenish blue Small grains and crystals with octahedral or dodecahedral form

274 JAHNS and EWING and many modifying faces, larger subhedral and roughly association with microlite. In part roughly cleavable. Rare euhedral crystals 0.2 to 0.8 in. across, and rare crystals re- mahogany-brown manganotantalite, associated with albite in portedly as large as 3 in. Chiefly as dispersed individuals in interior parts of main dike, forms 0.3- to 1.5 in. tablets with "spotted rock," in aggregates of lepidolite and Li-muscovite, distinct cleavage and bright, slightly resinous luster. Some and along margins of spodumene crystals; also as local dense crystals zoned, with thick rims of black tantalite. Very rare concentrations (Fig. 3). White, pale yellow to honey-yellow, bismutotantalite occurs as small, irregular residua in masses of yellowish to bronzy brown, dark maroon, and dark reddish earthy yellow bismutite found locally in coarse-grained quartz. brown to nearly black, with dull to vitreous luster. Many The 18 samples of Harding tantalite-columbite thus far crystals color zoned, generally with lighter cores and darker analyzed have ranged from Ta-rich columbite to nearly pure rims. Darkest material metamict or nearly so. tantalite, with weight ratios of Ta205 to Ta205+Nb205 Typical variations in composition among Harding microlites ranging from 0.45 to 0.94. The ratios for most crystals in the are shown in Table 4. With reference to the general formula hanging-wall zone of the main dike probably are near 0.6, and A2 B20 6X, A in these materials is chiefly Ca, Na, and U with those for crystals in the "spotted rock" near 0.8. very minor K, Mn, Fe, and Th. B is Ta and Nb with a little Ti, and X includes both OH and F. The range from light to dark PEGMATITE GENESIS colors can be correlated with progressive decreases in Ta+Nb content and Ta/Nb ratio, and with progressive increases in The Harding pegmatite bodies are best explained as OH/F ratio and U, Ti, and Mn content. The weight ratio of products of highly differentiated rest-liquid derived from granitic magma and injected into already deformed and Ta2O5 to Ta205+Nb205 is typically 0.86 to 0.89 for the darkest material, averages about 0.92 for the brownish yellow, regionally metamorphosed host rocks at considerable depth. and is 0.93 to 0.95 for the palest yellow to off-white material. An igneous origin, rather than metasomatic or metamorphic Of the 45 Harding specimens and bulk concentrates that have derivation from the host terrane, is indicated by several fea- been analyzed, all have proved to be tantalum-rich, and hence tures of occurrence. The bodies are distinctly discordant in to be microlite rather than pyrochlore. The uranian microlite, most places, for example, and they are sharply bounded from with a known maximum content of about 10 percent of the country rocks. Local severe deformation of these rocks UO2+UO3, commonly has been referred to the variety adjacent to margins of the bodies, clearly superimposed upon hatchettolite, even though it is not in the pyrochlore range. the results of earlier regional deformation, bespeaks forceful injection of pegmatite magma. Internal zoning of the dikes Tantalite-columbite [(Fe,Mn)(Ta,Nb)2O]6 . In outer parts of dikes as blade-like to thickly tabular black crystals with dull to reveals a consistent paragenetic sequence of major minerals, very bright luster. Most crystals 0.2 to 1.3 in. in maximum and the gross geometry of these zones reflects the shapes of dimension, but a few 3- to 4-in. waterworn chunks found in the containing bodies rather than features of the enclosing nearby placer deposits. Also scattered through "spotted rock" metamorphic rocks. The dikes themselves show no significant and other internal units as thick blades, plates, and equant compositional variations that could be correlated with lith- crystals 0.05 to 0.8 in. in maximum dimension, generally in ologically contrasting parts of the adjacent metamorphic terrane. Further, the pegmatites contain suites of Be, Li, Nb, Ta, U, and other trace elements at high concentration levels that are more compatible with an igneous than with a directly metamorphic derivation. The pegmatites are not obviously related, in either spatial or compositional senses, to any of the exposed granitic plutonites in the region. They are not systematically distributed with respect to the plutons, nor do they reflect the regional meta- morphism that left its stamp on most of the other rocks. If the remarkable abundances of rare elements in the Harding pegmatites are attributed to extreme magmatic differentiation, the compositions of known granitic plutonites of the region do not identify them as attractive candidates for a cogenetic relationship. The quartz monzonites in areas south and east of the Harding occurrences, for example, are very low-Ta, high-Ti rocks whereas the pegmatites are very rich in Ta and poor in Ti. The contrasting match-ups among these and other minor constituents are difficult to reconcile with any simple model of genetic association. Perhaps more important, the pegmatites may well be much younger than the youngest of the exposed granitic rocks (Long, 1974), and they may also represent a deeper level of emplacement. Like the pegmatites of the Petaca District west of the Rio Grande (Jahns, 1974), those in the Harding area probably were emplaced at depths of 7 miles or more, and were slowly cooled under high confining pressure in the presence of a halide-rich aqueous fluid. Emplacement was accomplished in a terrane of pervasively deformed rocks with relatively high metamorphic rank, and it occurred long after dacitic and other igneous rocks

were formed at relatively shallow depths. Many of the pegma- Circ. 7319,48 p. tite bodies are fringed by zones of metasomatic reaction with Berliner, M. H., 1949, Investigation of the Harding tantalum-lithium deposits, Taos County, N. Mex.: U.S. Bur. Mines Rept. Inv. 4607, 7 the wall rocks, and none of them contains crystal-lined P. pockets or primary cavities. Nor are they associated with Cameron, E. N., Jahns, R. H., Page, L. R., and McNair, A. H., 1949, aplitic rocks of the kind formed by sudden relief of confining Internal structure of granitic pegmatites: Econ. Geology Mon. 2, 115 P. pressure. The potent fluxing effect of volatile constituents Fullagar, P. D., and Shiver, W. S., 1973, Geochronology and petro- held in magmatic solution under high pressure is attested by chemistry of the Embudo Granite, New Mexico: Geol. Soc. America Bull., v. 84, p.2705-2712. subsolvus crystallization of alkali feldspars in the pegmatite Gresens, R. L., 1975, Geochronology of Precambrian metamorphic bodies, and especially by primary formation of nearly pure K rocks, north-central New Mexico: Geol. Soc. America Bull., v. 86, p. and Na phases in their central parts. 1444-1448. Heinrich, E. W., 1947, Beyerite from Colorado: Am. Mineralogist, v. Further discussion of the nature, gross segregation, and 32, p. 660-669. observed paragenetic relationships of the pegmatite minerals −--- and Levinson, A. A., 1953, Studies in the mica group; mineralogy of rose muscovites: Am. Mineralogist, v. 38, p. 25-49. cannot be offered here, but in sum these features are com- Hess, F. L., 1933, Pegmatites: Econ. Geology, v. 28, p. 447-462. patible with the genetic model for pegmatite crystallization Hirschi, H., 1928, Thormineral aus lithiumpegmatit von camp Harding outlined by Jahns and Burnham (1969) on the basis of field, bei Embudo, New Mexico: Schweizerische Mineralog. und Petro- graph. Mitteil., v. 8, p. 260-261. petrographic, and experimental investigations. Most of the −--- 1931, Mikrolith in spodumenpegmatit bei Embudo in New crystallization probably occurred in the presence of both Mexico: Schweizerische Mineralog. und Petrograph. Mitteil., v. 11, p. silicate melt and dense aqueous vapor rich in F and Cl, and 253-255. Jahns, R. H., 1951, Geology, mining, and uses of strategic pegmatites: nearly all the remainder in the presence of the vapor. Exis- Amer. Inst. Mining, Met. and Petrol. Engineers Trans., v. 190, p. tence of this vapor can be inferred on several grounds, and is 45-59. more directly attested by numerous fluid inclusions in the ---- 1953a, The genesis of pegmatites: I. Occurrence and origin of giant crystals: Am. Mineralogist, v. 38, p. 563-598. pegmatite quartz and beryl. Its influences as a catalyst and as a − 1953b, The genesis of pegmatites: II. Quantitative analysis of medium for material transfer are implicit in the metasomatic lithium-bearing pegmatites, Mora County, New Mexico: Am. Mineral- ogist, v. 38, p. 1078-1112. interactions between dikes and country rocks, in the nearly −--- 1974, Structural and petrogenetic relationships of pegmatites in complete unmixing of early-formed alkali feldspars to yield the Petaca district, New Mexico: New Mexico Geol. Soc. Guidebook, coarse perthite, in the leaching of calcium from the feldspars 25th Field Conf., Ghost Ranch (Central-Northern New Mexico), p. for fixing in microlite and late-stage apatite, and in widespread 371-375. ---- and Burnham, C. W., 1969, Experimental studies of pegmatite mineral recystallization and replacement under both hyper- genesis: I. A model for the derivation and crystallization of granitic solidus and subsolidus conditions. The most extensive of the pegmatites: Econ. Geology, v. 64, P. 843-864. replacement reactions led to formation of albite at the expense Johnson, J. H., 1940, Iceland spar mine in New Mexico, with notes on the properties and uses of the spar: Mines Mag., v. 30, P. 59-60. of quartz and microcline, and micas at the expense of micro- Just, Evan, 1937, Geology and economic features of the pegmatites of cline and spodumene. Taos and Rio Arriba Counties, New Mexico: New Mexico Bureau Mines & Mineral Resources Bull. 13,73 P. REFERENCES Aldrich, L. T., Wetherill, G. W., Davis, G. L., and Tilton, G. R., 1958, Radioactive ages of micas from granitic rocks by Rb-Sr and K-Ar methods: Amer. Geophys. Union Trans., v. 39, p. 1124-1134. Baker, J. S., 1945, World survey of tantalum ore: U.S. Bur. Mines. Inf. 276 JAHNS and EWING Kelley, V. C., 1940, Iceland spar in New Mexico: Am. Mineralogist, V. Northrop, S. A., 1942, Minerals of New Mexico: University of New 25, p. 357-367. Mexico Bull. 379, Geol. ser., v. 6, 387 p.; 1959 (revised edition): Levinson, A. A., 1953, Studies in the mica group; relationship between Albuquerque, University of New Mexico Press, 66s p. polymorphism and composition in the muscovite-lepidolite series: Page, L. R., 1950, Uranium in pegmatites: Econ. Geology, v. 45, p. Am. Mineralogist, v. 38, p. 88-107. 12-34. Long, P. E., 1974, Contrasting types of Precambrian granitic rocks in Rimal, D. N., 1962, Mineralogy of rose muscovite and lepidolite from the Dixon-Penasco area, northern New Mexico: New Mexico Geol. the Harding pegmatite, Taos County, New Mexico [Ph.D. dissertl Soc. Guidebook, 25th Field Conf., Ghost Ranch (Central-Northern Albuquerque, University of New Mexico, 94 p. New Mexico), p. 101-108. Roos, Alford, 1926, Mining lepidolite in New Mexico: Eng. and Mining Montgomery, A., 1950, Geochemistry of tantalum in the Harding Jour., v. 121, p. 1037-1042. pegmatite, Taos County, New Mexico: Am. Mineralogist, v. 35, p. Schaller, W. T., and Henderson, E. P., 1926, Purple muscovite from 853-866. New Mexico: Am. Mineralogist, v. 11, p. 5-16. 1951, The Harding pegmatite—remarkable storehouse of massive ----, Stevens, R. E., and Jahns, R. H., 1962, An unusual beryl from white beryl: Mining World, v. 13, p. 32-35. Arizona: Am. Mineralogist, v. 47, p. 672-699. ---- 1953, Pre-Cambrian geology of the Picuris Range, north-central Soule, J. H., 1946, Exploration of Harding tantalum-lithium deposits, New Mexico: New Mexico Bureau Mines & Mineral Resources Bull. Taos County, N. Mex.: U.S. Bur. Mines Rept. Inv. 3986, 10 p. 30, 89 p. Wood, J. A., 1946, Spiral concentrator recovers tantalum ores at the Mrose, M. E., 1952, The a-eucryptite problem [abs.]: Geol. Soc. Amer- Harding mine: Mining Cong. Jour., v. 32, p. 44-45. ica Bull., v. 63, p. 1283.